# deepseek-ai/open-infra-index

Production-tested AI infrastructure tools for efficient AGI development and community-driven innovation

Repository: https://github.com/deepseek-ai/open-infra-index
Canonical: https://ross.abutalabs.com/products/open-infra-index
License: CC0-1.0
License Family: permissive
Last push: 2025-05-15T02:00:43+00:00

## Health v2 (maintenance only)
Score: 30/100 (v2, computed 2026-09-03T02:39:23.370411+00:00)
- activity 21, release rhythm 35, longevity 39
- inputs: {"age_days": 559, "days_push": 476, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: no_releases
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 8059, forks 298 (observed 2026-08-28T04:10:12.367528+00:00)

## What it is
An index repository from DeepSeek AI cataloging their production-tested open-source AI infrastructure tools, including FlashMLA, DeepEP, DeepGEMM, and parallelism strategies. It serves as a hub linking to GPU kernel libraries and communication libraries used in DeepSeek-V3/R1 training and inference.

## Use cases
- find DeepSeek's open-source GPU kernels for LLM inference
- learn how DeepSeek scaled V3 training on Hopper GPUs
- discover MoE communication libraries like DeepEP
- find FP8 GEMM libraries for H800 GPUs
- study production AI infrastructure from DeepSeek's open-source week
- locate pipeline parallelism algorithms like DualPipe

## When to avoid
- you need a runnable tool rather than an index of links
- you don't work with NVIDIA Hopper-class GPUs or large-scale LLM infrastructure

## Facets
- artifact type: learning-resource
- maturity: active
- function: gpu-computing, llm-inference, llm-training, developer-tools
- domain: artificial-intelligence, large-language-models, gpu-computing, infrastructure-as-code, developer-tools
- platform: cpp, python
- tags: deepseek, moe, fp8, hopper-gpus, index-repo, open-source-week, inference-engine, parallelism, gpu, linux

## Member repositories
- deepseek-ai/open-infra-index (main) score 30

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:10:12.367528+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-29T17:31:23.103595+00:00, confidence not recorded.
  - readme: https://github.com/deepseek-ai/open-infra-index (fetched 2026-08-28T04:10:12.367528+00:00, sha aba10361316b)
- Data as of 2026-08-30T08:39:29.467469+00:00.